The long-term objective of this project is to understand the role of cortical oculomotor centers in programming and initiating voluntary saccades. The planned experiments of the present project period concern particularly the neural correlates of antisaccades and triple-step saccades performed by monkeys. Instead of being directed to visible targets, antisaccades must be aimed in exactly the opposite direction. Thus, the location of their goals has to be internally computed. This represents a high level of performance in which patients with frontal lobe lesions are severe deficient: they can hardly refrain from reflexively glancing at the target itself before making the required antisaccade. We have already trained monkeys to perform antisaccades and recorded neural correlates of their successes and failures in this task. Microelectrode recording from single, isolated nerve cells and microstimulation will be performed in macaques trained to perform antisaccades. The experimental paradigm includes random alternance of pro- and antisaccades upon instruction given by the shape of a central fixation point and random alternance of the location of peripheral stimuli indication the vector of pro- or antisaccades to be executed. The triple-step task was conceived for and used in the past period of the grant for distinguishing the role of tonic visual cells in mapping the location of targets stored in memory from a role in preparing the next saccades. The paradigm consists in presenting three flashed targets in close succession before the eyes move from a fixation point to the location of the first target. Use of this task will also bring complementary information about the mechanism of antisaccade generation. The investigated cortical areas are the supplementary eye field, the frontal eye field-both in the frontal lobe-, and the lateral intraparietal cortex. In some experiments, microstimulation will be performed during the performance of both tasks, to probe the system under study at specific interval in order to determine the timing of the events disturbed by this method.